Microsoft Focuses on Azure and AI during Annual Dev Build

Image of a phone with Microsoft products showcased on 2022 Dev Build Conference.
This year’s Dev Build conference was focused on AI, Cloud, and Mobile.

Despite the lasting COVID-19 pandemic, Microsoft did not cancel the 2022 Dev Build conference. It took place from the 24th to the 26th of May, completely online.

During the conference, the company disclosed several new projects and tools. These aim to help devs improve their efficiency and focus on new products with more assistance from hardware and software. The conference primarily focused on Microsoft Azure and new AI developments. It also discussed the hardware that will support these advancements.

The Dev Build conference discussed these 7 key projects:

  1. Microsoft Dev Box
  2. The Responsible AI dashboard
  3. Updated Azure Cognitive Services (w/ Open AI)
  4. Partnership with Meta (formerly Facebook)
  5. Partnership with AMD
  6. New Project Volterra
  7. Compatibility with the Hugging Platform

Noticeably, the conference did not focus on Microsoft’s cloud platform and AI development tools as separate subjects. Rather, they were brought up as parts of a whole. New tools will integrate with the cloud and machine learning software. As such, they could predict and solve mistakes before they even form. Developers also believe completely new solutions might be possible soon.

In this Dev Build conference, Microsoft focused on developers more than acquisitions and strategies. Hopefully, this will make Azure DevOps even easier to use.

Importance of Cloud Computing

For most businesses at the moment, the cloud is used almost exclusively for backups and disaster recovery solutions. But, with the development of new tools, it will be possible to use virtual machines for even more purposes. As such, we could push entire business models to the cloud.

This move will strongly impact companies focusing on remote work. In these situations, you would never need anyone to be in a fixed location. Rather, your employees could access their tools and resources from any machine.

Switching to the cloud will also move hardware resource demands away from individual workstations. Instead, we will see a collective use paradigm. In turn, that significantly reduces the amortization and usage costs of hardware for companies. Even cloud cost optimization will become a necessary part of any model.

Illustration of server symbols overlaying a server room.
In the future, more than just data will be on the cloud.

Dev Build Introduces New Microsoft Dev Box

At the Dev Build Conference, Microsoft showcased a new service dubbed the Dev Box. It focuses specifically on developers. This service will build on the Azure Virtual Desktop. Design-wise, it is similar to Windows 365 but with a different emphasis.

The new service will allow developers to access pre-configured virtual machine systems. It will use the Software Development Kit (SDK), the Integrated Development Environment (IDE), or any other development tools based on top of a Windows machine.

Developers were most interested in the debug option. Namely, by using the Microsoft Dev Box, it will be possible to debug working code without debugging the device you are working on.  The waiting time on the rebuilds will also be significantly reduced.

According to Anthony Cangialosi, a group program manager at Microsoft, “the new service will ensure compliance, unified management, and increased security. It will use Win365 to integrate the Dev Box with MEM (Microsoft Endpoint Manager) and Intune.”

This will make advanced AI easier to use and code for developers. That’s especially true for the responsible AI dashboard features. The AI dashboard will combine analytics, model interpretation, and other AI features to allow new options for future developers. It will also make them faster and easier to use.

New Project Volterra Tools

The 2022 Dev Build conference has given Microsoft a chance to increase support for ARM CPUs. Most notably, Microsoft can now develop AI and Neural Processing Units, or NPUs. These are specialized cores that should increase the AI learning speed. This will greatly improve machine learning algorithms.

In that direction, the company has introduced what they are calling Project Volterra. This is a compact PC build that will be using the Snapdragon SoC (System on Chip), opting against anything made by the two CPU giants, AMD or Intel.

By using the new build with the Qualcomm Neural Processing SDK built exclusively for Windows, developers will be able to test multiple AI scenarios at a more accessible rate.

In the words of Panos Panay, the CPO for Microsoft, “the company will produce a myriad of options for ARM development tools that will be utilized by Project Volterra and its users.” These tools include:

  • Full Visual Studio 2022
  • VSCode
  • Visual C++
  • Modern .NET 6 Java
  • Classic .NET Framework
  • Windows Terminal
  • WSL
  • WSA

Combined, the new Volterra project tools will help test AI solutions for multiple platforms. Developers believe that it will greatly assist app development for Android and web services.

Image of the AMD Ryzen Chip in a Motherboard.

Powered by AMD MI200

During the Dev Build conference, Microsoft announced that the AMD MI 200 GPU will be used for AI training on Azure. According to the company, this is the first time any public cloud service uses this hardware. It is also reportedly five times faster in computing power than the Nvidia A100, which is currently used in Azure-based AI clusters.

The MI 200 is a Dual-GPU chiplet with 96 TFLOPS that builds on AMD’s CDNA2 architecture. During the last 6 months that it has been available for the public, the improvements over the MI100 became apparent. In fact, its dual-die architecture increased computer performance.

Additionally, with a total of 58 billion transistors, the MI200 is very compact. On a whole GPU, it uses what the A100 uses on a single core. In theory, that will allow the use of a lot more serial GPUs without an increase in area, power, and cooling requirements.

Leave a Comment

Your email address will not be published.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Scroll to Top